Microsoft snafu calls into question its cloud reliability

You’d think Microsoft could at least do a decent job of running a Web site. But its new Volume Licensing Service Center malfunctioned for much of December, leaving resellers and their customers in the cold. The foul-up, and the company’s tepid and belated response to angry customers, belies claims that Microsoft has put its Vista-era troubles behind and raises new questions about the reliability of its cloud-based services.

What happened was a bit complicated, but the ramifications were pretty simple: Resellers, integrators, and enterprise customers rely on the VLSC to track and update license information, download software, and so on. Authorizing a new user, for example, requires getting a key from the site — no site, no key, no access.

[ Are your software licenses keeping up with the times? Probably not — and that can cost you big time. Here’s how to win the licensing game. | You can expect more cloud computing outages in 2010, predicts InfoWorld’s David Linthicum. ]

Previously, resellers went to Microsoft’s eOpen and MVLS sites for that information. But when Microsoft replaced them with the VLSC on Dec. 9, users found that the new site was down. And as one angry reseller told me, “When the system came back up, the license information that was in the old system had vanished. Microsoft has made the license purchase/fulfillment process a multistep nightmare.”

The net result: wasted time and lost business for the very companies that Microsoft relies on to sell and service enterprise products. And who knows how many hassles for the rank-and-file IT folks who had to explain why their users couldn’t access needed applications.

As of this week, the site appears to be functioning, though there are still some reports of problems.

Microsoft outage a lesson in how to do it wrong

We all know that stuff happens: Web sites go down, glitches pop up, software and hardware don’t always work. But companies that care about their customers are quick to fess up and let users know when the problem will be resolved. That’s not how Microsoft handled the VLSC outage.

On Dec. 14, nearly a week after the new site should have been up and running — but wasn’t — Richard Gibbons, a software manager with Bechtle UK, posted this comment: “From my point of view as a Microsoft licensing reseller, I haven’t heard or seen anything from Microsoft on this subject at all. No heads-up so we can warn our customers, or at least be prepared for the questions, no advice on how we can help, etc.” Along with his day job, Gibbons hosts a blog that became a point of contact for users during the outage.

In an e-mail to me, Gibbons adds, “Even as one of the top partners in the U.K., I was unable to get much more information internally, and the Web site gave only a generic message with the usual contact details. I asked Microsoft’s PR company Waggener Edstrom for an update/statement on December 14th and have so far received nothing.

[Update: Microsoft statement on outage at end of post]

Because resellers weren’t getting information from Microsoft, they couldn’t tell their customers how to proceed. A reseller and systems integrator who posts under the name “Duke” had a customer who wanted a license for Access. Duke directed him to the Microsoft site, thinking it was back in operation. It wasn’t. He writes, “Our customer is hopping mad. Any customers considering using Microsoft’s cloud services take note, not just of the failure but of Microsoft’s inability (or lack of concern) to keep customers apprised of the situation.”

To make matter even worse, there was another, apparently related issue that caused problems too, says Gibbons. “The outage appears to have also affected a system called MOET, which distributors use to place certain licensing orders to Microsoft. This caused two- to three-week delays in orders being processed.” And it didn’t help at all that all this occurred in December, as companies closed out their budgets, he adds.

A shadow on the cloud After customers were told the site would be up by Dec. 16, the date slipped two more days, and when it was finally running, the woes continued. “Jon” posted this: “Yes it is back up, but the problem remains. No way to add a new agreement, and the agreement I added over the phone doesn’t show up either. Two full weeks of ‘maintenance’ yields nothing. We are right back where we started. Trash this project and bring back eOpen. It sucked, but it worked. This VLSC is just stupid. Stop it already.”

One bright spot: Microsoft’s techies used Twitter rather effectively to help some customers. And it’s likely that the VLSC will be an improvement over the former clunky system — once the bugs are really eliminated.

Other than those slightly saving graces, this whole episode stinks. It’s bad enough that Microsoft was too quick to launch an important project without adequate testing, but leaving its customers hanging for so long is inexcusable.

The VLSC is not really a cloud service. But a number of posters on the Gibbons site note that it makes them wary of Microsoft’s nascent cloud ambitions. “BurnandCandle” put it well, saying, “If this is the kind of service and reliability we can expect from the cloud, then count me out. The users that I support won’t meekly accept some lame Web page apology instead of immediate access to their resources.”

Update: Two and a half days after I asked  for Microsoft’s response to the VLSC situation, I got the answer pasted below.  It’s worth noting that while an apology is always welcome, there’s no indication of when the problems will be over and what actually caused the issue. Saying there was an “issue with the registration system” doesn’t explain anything anything. Meanwhile, I’m still getting notes from users who can’t access the system.

“We apologize to our partners and customers for the inconvenience that they have been through during the upgrade improvements for the Volume Licensing Service Center (VLSC) site. In standard systems testing, we encountered an issue with the registration system.  While the vast majority of partners and customers are able to access the system, there remain some issues that are causing difficulties for some and it has taken us longer than expected to correct these issues. We are taking all necessary steps to resolve the situation and we are working with each impacted customer or partner to restore permissions if they can’t be resolved online. We understand the inconvenience that this causes and greatly value doing business with our partners and customers.”

I welcome your comments, tips, and suggestions. Post them here so that all our readers can share them, or reach me at bill.snyder@sbcglobal.net.

This article, “Microsoft snafu calls into question its cloud reliability,” was originally published at InfoWorld.com. Follow the latest developments on Microsoft and cloud computing at InfoWorld.com.